At GI-KACE, I co-developed two internal systems using Python FastAPI and PostgreSQL, replacing paper attendance sheets and manual staff appraisals with structured digital records. I also tested and validated backend APIs in Swagger.
In my research, I cleaned cross-country time-series data and compared statistical and machine learning forecasts using RMSE. I also built the data cleaning and preprocessing pipeline for an early-warning model for floods and droughts, and co-authored the research paper.
On a project called Stateful Multi-Agent IT Support System, I built an automated evaluation framework to track groundedness, tool-selection accuracy, task completion, and hallucination rate, then used the metrics to decide what to fix next. I'm currently designing Agentic BI: Ask Your Data, a prototype for asking questions about company financial data in plain language.
I'm completing a Master's in Artificial Intelligence at Carnegie Mellon University Africa. I've also co-taught Python, Office Productivity, and Data Analytics, and bring experience working with distributed teams.
